文章基本信息

标题：A NEURO COMPUTING FRAME WORK FOR THYROID DISEASE DIAGNOSIS USING MACHINE LEARNING TECHNIQUES
本地全文：下载
作者：SHAIK RAZIA ; M.R.NARASINGARAO
期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2017
卷号：95
期号：9
出版社：Journal of Theoretical and Applied
摘要：Thyroid is a disease, managing of which a difficult proposition in a clinical set up. Many researchers have developed different machine learning techniques to diagnose the disease. Different models had been developed using different algorithms for prediction of the disease. In this proposed research, an integrated model using two neuronal models (SOM and LVQ) have been developed for the purpose. The unlabeled dataset consisting of 215 instances has been gathered from UCI repository. Five inputs have been considered for the model. They are Triiodothyronine, serum Thyroxin, Total Serum Triiodothyronine, TSH, Max TSH and the output is considered as status which has values 1(hyper),2(hypo) and 3(Normal). The integrated model has been developed using competitive learning algorithm along with vector quantization algorithm. The outcome of the integrated model has been compared with decision tree model in predicting the disease and it is observed that the integrated model outperformed the decision tree model with respect to the accuracy of the network. The integrated model has been developed with .net technology and the decision tree model is generated using statistical programming language using R.
关键词：Self Organized Maps; Linear Vector Quantization; Triiodothyronine; Serum Thyroxin; Serum Triiodothyronine; Thyroid Stimulating Hormone; Hyper Thyroid; Hypothyroid; Decision Tree.